Superconductors with Superconducting Transition Temperatures Tc = 91K (1999), 120K (1994), 340K (2000), and 371K (1995): Experimental Errors or a Technological Puzzle? Two-Component Nonstoichiometric Compounds and the Insulator--Superconductor--Metal Transition

V.N. Bogomolov
DOI: https://doi.org/10.48550/arXiv.cond-mat/0311265
2003-11-17
Abstract:One of the reasons for the lack of understanding of both the mechanisms underlying the HTSC phenomenon and of the instability of materials with Tc > 300 K may be the widely accepted but wrong ideas about the types of chemical bonding in a substance and the radii of the atoms and ions. A revision of these concepts started in the beginning of the XX century in connection with the investigation of non-stoichiometric compounds (the berthollides) but did not reach a critical level until recently. Most of the HTSC materials, however, are actually non-stoichiometric nanocomposites, whose components "dilute" or "stretch" one another. Each component resides in an "intermediate" state, which still remains poorly studied. For instance, in a system of particles having two paired electrons each, the unbroken electron pairs may start tunneling at a certain "medium" concentration with the system becoming a Bose superconductor (the state between the insulator and the metal with BCS superconductivity). For univalent atoms (Na,Ag), however, such possibility realizes neither in intermediate nor in the final state. Univalent metals are not superconductors. In the berthollides, however, a possible Jahn-Teller-Peierls-type instability may give rise to formation of diatomic molecules (Na2, Ag2)with electron pairs, and superconductivity can set in. It is possibly such systems that were obtained by chance in experiments with univalent components and reported to have Tc of up to 371 K. Structures of a number of HTSC materials are considered.
Superconductivity
What problem does this paper attempt to address?